[sinae]

117DB crosstalk is very bad in my opinion.. like if crosstalk isnt important in a dac.. it's not truly dual mono..

who cares about -125db of SNR if you have -117 DB crosstalk..

i personnally find funny that they sell the yddrasil at this price.. i would go audio-dg R28 for half the cost in a heartbeat

 

[hornytoad]

117DB crosstalk is very bad in my opinion.. like if crosstalk isnt important in a dac.. it's not truly dual mono..

who cares about -125db of SNR if you have -117 DB crosstalk..

i personnally find funny that they sell the yddrasil at this price.. i would go audio-dg R28 for half the cost in a heartbeat

Can you hear the crosstalk ?

 

[DrTone]

117DB crosstalk is very bad in my opinion.. like if crosstalk isnt important in a dac.. it's not truly dual mono..

who cares about -125db of SNR if you have -117 DB crosstalk..

i personnally find funny that they sell the yddrasil at this price.. i would go audio-dg R28 for half the cost in a heartbeat

I'm fine with -146db crosstalk with the balanced out.

 

[adydula]

Somehow I dont think you know what crosstalk is, how to measure and what it really means....

Alex

 

[astromo]

@jude thanks for being open and detailed with your methodology.

With my (non electrical) engineering background, what goes into good science is transparency of methodology so that anyone else can repeat a piece of empirical work to enable reliable comparison. Another part of good science is peer or expert review.

The guy that you're putting up to the technical challenge obfuscates his methodology (as I've read it) and doesn't appear to engage with the instrument vendor for tech support (if he does, then he's very quiet about it).

So, I can see that your work is striving to achieve good science. The work that you're comparing with, falls seriously short of this goal (to my eyes at least).

Thanks for all the effort.

 

[adydula]

Thanks for the update....I honestly do not care anymore about what or anything amirim is measuring or doing. Period.
Used to think he was just trying to make things better overall but after all the dust has settled it obvious that I can not trust what
he is doing...

My ears tell me what is good or bad in the final decision....and so far the Schitt stuff I have, and have had has been nothing but good working stuff at a fair price IMO, made in the USA!

Looks like the Yggdrasil specs posted above are pretty good to me.

Alex

 

[astromo]

@jude thanks for being open and detailed with your methodology.

With my (non electrical) engineering background, what goes into good science is transparency of methodology so that anyone else can repeat a piece of empirical work to enable reliable comparison. Another part of good science is peer or expert review.

The guy that you're putting up to the technical challenge obfuscates his methodology (as I've read it) and doesn't appear to engage with the instrument vendor for tech support (if he does, then he's very quiet about it).

So, I can see that your work is striving to achieve good science. The work that you're comparing with, falls seriously short of this goal (to my eyes at least).

Thanks for all the effort.

For info, I've been sent a PM regarding my post by @amirm

He claims to have been locked out of this forum and hence cannot respond to this thread. I advised that he should take the issue up with @jude directly and present his case.

I provided further detail to @amirm to justify my statements here. In short, because his results appear to be at variance with those presented by others I put it to him that if he wants acceptance that his results are valid and correct, then he'll have to do further testwork and part of that is being clear and open with test methodology. Effectively he's the salmon swimming up the stream of alternate testwork results. That's a challenging prospect. Especially for @amirm where he's been exiled from various online communities and has admitted to me that he has difficulties in accessing an Yggdrasil unit to do further testing. I put it to him that to address a number of validity questions, he really needs to repeat his testwork with his new APx555 that is undoubtedly within its calibration period. Referring back to results from an old instrument that's out of test will always attract doubt because in my mind that's not a basis for results that are "100% valid".

@amirm claimed that @jude has not shared AP test protocol files with him. I simply responded by challenging him to publish his test files on a public site and set the bar for @jude to meet (and anyone else who has an AP instrument and has a desire to repeat the work). As I've stated, part of good science is publishing your method clearly so that anyone else with the resources and wherewithal can repeat your work and, if everything's been done right and is valid, arrive at the same results.

Note, I've purposely not simply done a copy/pasta of the PM with @amirm so you'll have to trust me on what I've related above. Given that the message is "private", I think it's appropriate to keep the verbatim detail off the grid. If he so desires, I'm happy for him to publish the full detail over at his own website given that he appears not to be able to do that here.

 

[Alcophone]

Especially for @amirm where he's been exiled from various online communities and has admitted to me that he has difficulties in accessing an Yggdrasil unit to do further testing.

Well, he received mine this morning, and I told him he can keep it for up to three weeks.

@amirm claimed that @jude has not shared AP test protocol files with him. I simply responded by challenging him to publish his test files on a public site and set the bar for @jude to meet (and anyone else who has an AP instrument and has a desire to repeat the work).

There was some discussion on his site about that, and one of his reasons not to publish it was that it's apparently quite a big file (50 MB or so?) and that the hosting costs are therefore noteworthy. But at least one member offered him to take care of that, while others proposed additional solutions. The thread on his site is currently locked, though.

I've purposely not simply done a copy/pasta of the PM with @amirm so you'll have to trust me on what I've related above.

Well, even if it looked like you copied & pasted, we'd have to trust you not to have altered it.
In general, I think it's good not to publish PMs without permission, so kudos for that.

 

[astromo]

Well, he received mine this morning, and I told him he can keep it for up to three weeks.

@Alcophone that rates some serious kudos points right there. s off to you.

There was some discussion on his site about that, and one of his reasons not to publish it was that it's apparently quite a big file (50 MB or so?) and that the hosting costs are therefore noteworthy.

Seriously? I would have thought that this would be pretty straight forward in today's world for an ex M$ VP. I'm not blessed with such credentials just some nifty fingers that got me to:
https://www.howtogeek.com/howto/36947/how-to-upload-really-large-files-to-skydrive-dropbox-or-email/
As Oscar Goldman said it, " ... we have the technology".

 

[adydula]

I read his locked thread. He was asked by AP as well for his files, and he refused according to him in this thread.

IMO, this subject should be over.

This guy keeps shooting himself in the foot and after awhile you loose all respect and credibility.

Let it go...move on...and listen to you Schitt…

Alex

 

[phase0]

There was some discussion on his site about that, and one of his reasons not to publish it was that it's apparently quite a big file (50 MB or so?) and that the hosting costs are therefore noteworthy. But at least one member offered him to take care of that, while others proposed additional solutions. The thread on his site is currently locked, though.

That's a joke. Obvious dodge. Anyone try to share a 50MB file over the internet in the last 10 years? I can assure you it's definitely possible.

Edit:

I loaded his front page and saved it to disk using a web browser. That came out to 628 KB. 1024 KB = 1 MB.

How many visitors just loading the front page would equal 50MB? 51,200 / 628 KB = around 81.5. So just loading his front page of his site 82 times will cause 50 MB of traffic on his site.

How many visitors does his site get? I looked at similarweb.com. According to them ASR received 200,000 visitors in April 2018, and 180,000 in May 2018.

So if all those 180000 hits did was load the front page at 628KB and did nothing else at all on the site (no pictures of equip etc, no browsing sub pages/reading/posting), that would be 180000 * 628 KB / 1024 (KB per MB) == 110,390 MB for one month.

How much is a 50 GB file represent of that much traffic? 0.05 % of their monthly traffic. (Again if all the visitors did is just load the front page).

So really?! Bandwidth to host this config file is just going to push that monthly bill over the edge? That site must be pushing way more traffic than my little conservative estimate.

 

[atomicbob]

Quibble over 50MB? Seriously??!! That is hardly 3 minutes of a single high resolution track (96x24) in a musical collaboration sent between studios. Larger files are exchanged all the time. Usually multiple tracks too.

 

[Alcophone]

I should clarify that the discussion was about publicly sharing the file, i.e. 50 MB per download. Personally, I would think twice before I casually hosted that on my website, because that can add up quickly.
Of course where there's a will, there's a way. But he says he's still busy replicating his old measurements with his new analyzer, so his files aren't final yet anyway. So it's understandable that he's not eager to share them broadly just yet.

When it comes to sharing them with just Jude or AP, I agree, I would have simply done that in the interest of a transparent discourse as well as reproducibility (pretty important part of science). I could see that he would not want Jude to have his email address or something, but as has been said here and on his site, there are many other ways to share files of that size conveniently with a select group of people.

Completely apart from that: why are those files so big? That really surprised me. How compressible are they?

 

[phase0]

should clarify that the discussion was about publicly sharing the file, i.e. 50 MB per download. Personally, I would think twice before I casually hosted that on my website, because that can add up quickly.

Just to pick a random example for hosting: https://www.homepageuniverse.com/

If you went there to set up your website you'll note that they now advertise "Unmetered Bandwidth" and "Unlimited Bandwidth". It used to be bandwidth cost a lot. Looking all over at a bunch of hosting providers tonight I am having a really difficult time finding any hosting that lists metering costs for network traffic any more. Now it's more about guaranteed throughput with a set amount of burst capacity from what I can see. It looks like the days of N GB per month metered bandwidth are gone (unless you're a home internet user of course because the Monopoly ISPs can bleed you dry).

 

[Alcophone]

Just to pick a random example for hosting: https://www.homepageuniverse.com/

If you went there to set up your website you'll note that they now advertise "Unmetered Bandwidth" and "Unlimited Bandwidth". It used to be bandwidth cost a lot. Looking all over at a bunch of hosting providers tonight I am having a really difficult time finding any hosting that lists metering costs for network traffic any more. Now it's more about guaranteed throughput with a set amount of burst capacity from what I can see. It looks like the days of N GB per month metered bandwidth are gone (unless you're a home internet user of course because the Monopoly ISPs can bleed you dry).

Probably true. I remember running into a limit of 10 GB/month with a site I hosted, but that was over 10 years ago. I'm sure if he had a burning desire to share the file publicly, he'd find a way.

I didn't mean to start a discussion about how to cheaply host larger files, though. That's off-topic. I only wanted to relay the proclaimed reason.

I still have to digest Jude's response. Already learned some more because of it, which I like!
I somehow thought the APx555 has a noise floor more than 140 dB down, but evidently not. So my theory about Amir's optimizations is likely moot. Well, no surprise there ^^